AVS Forum banner

Status
Not open for further replies.
1 - 20 of 173 Posts

·
Registered
Joined
·
412 Posts
Discussion Starter #1
On a whim while considering the plight of our friends here paying long distance for nightly downloads and using up their line while they have broadband access, I thought of an idea: why not do the following:


1. configure your computer (Win98, 2k, and Linux support this I believe) as a dialup server which hands out a NAT address to incoming call on your now-gathering-dust modem. Make sure there is no username or password required. Have it auto-reset the PPP connection (but not the modem) on 10 minutes idle.


2. configure the computer modem to connect without requiring a ring voltage (can't remember from my BBS days if this entails using the &L1 leased line parameter, ATA, ATH1 for off-hook, or somthing else).


3. Hack the Replay to change the modem init string to disable dialtone detection (X0).


4. Connect the computer and Replay *directly* together (i.e., not through the phone line).


Now, the Replay dials the computer (the computer ignores the dial tones), they negotiate, and the Replay's traffic is routed through the broadband connection.


If rings and dialtones are needed to make this work right, or if hacking the modem init string is too difficult, you could also use a phone line simulator. Viking Electronics ( www.vikingelectronics.com ) makes one that may do the trick (no pricing on their web site, just found it on a web search). You could also modify a circuit like the one at http://www.hotspot.freeserve.co.uk/H...S/tlsframe.htm (intended for UK POTS, not US).
 

·
Registered
Joined
·
1,595 Posts
Good idea. I would like to see if someone can make this work. However, doing it without hacking the Replay would be nice.
 

·
Registered
Joined
·
1,708 Posts
Just a bump for this thread in case anyone technical would be able to do this. This would be absolutely awesome it would allow me to dump my local phone company (the only reason I have one is b/c of the replay) and rely solely on my cell phone.



Scyber
 

·
Registered
Joined
·
99 Posts
Quote:
Originally posted by richardtallent:
3. Hack the Replay to change the modem init string to disable dialtone detection (X0).
No need to "hack" the modem init string... pressing ZONES while viewing the Dialing Prefix screen during Setup gets you to the undocumented Modem Settings screen which allows you to set Wait for dial tone (Yes/No), and Dial using (Tone/Pulse).


Ben.

 

·
Registered
Joined
·
65 Posts
ReplayBen is COOL!


Given that this process would eliminate dial-up connect fees, I can understand why somebody "in the know" would give us a hint.


Thanks!


Ed
 

·
Registered
Joined
·
1,951 Posts
I don't have anything new to add to the technical details, but if this is accomplished I can get rid of my home phone line.


Cheers


------------------

~~~~~~~~~~~~~~~~~~~~~~~~~

ReplayTV -- The only one that had firewire...
 

·
Registered
Joined
·
1,557 Posts
Quote:
Originally posted by ReplayBen:
No need to "hack" the modem init string... pressing ZONES while viewing the Dialing Prefix screen during Setup gets you to the undocumented Modem Settings screen which allows you to set Wait for dial tone (Yes/No), and Dial using (Tone/Pulse).


Ben.

Thank you, Ben. I now hate our phone company voice mail system much less than I used to.



------------------

PRMan
 

·
Registered
Joined
·
585 Posts
When you connect modem to modem, I don't think you can get 56K connections, can you? I think you are limited to 33.6K.


Not that this is a big deal, just curious.


Space
 

·
Registered
Joined
·
3,686 Posts
Pull this one off, Sean, and post instructions here and you will be well-loved indeed.


Still feel like a schmuck letting those boxes dial in every night when I have a perfectly good cablemodem.


'Course, that would mean having to leave a 'pute powered up to handle the PPP connection...


If only someone made a PPP/modem server that would plug directly into an ethernet hub... Yeah, I know, someone probably does, and it probably costs more than a ReplayTV.
 

·
Registered
Joined
·
339 Posts
Quote:
Originally posted by seanriddle:
I'm not sure- part of the 56K issue is the number of D/A converters in the connection, and there would be zero in this case. I had a USR ISDN modem that I could connect to with a 56K modem at >33.6K, but I'm not sure if 2 standard modems will or not. I guess I'll find out http://www.avsforum.com/ubb/smile.gif
Two pc modems will not connect to each other at 56K. Your standard 56K modem can only receive 56K and send 33K. So talking to each other will be constrained by both sides send limit which is 33K.

 

·
Registered
Joined
·
3,715 Posts
Anyone making any progress with this yet?

Just curious.


-Gary
 

·
Registered
Joined
·
75 Posts
Ok - So I can make my Replay dial my Linux box over my second phone line without a whole lot of work. Setting up Linux to authenticate and serve an IP was not too bad either. TACACS authentication helped with the account name issues so I didn't have to hack the Replay - probably a better solution there somewhere, but I just wanted to get this portion working.


Now - let's get the damn phone lines out of the mix. Anyone out there with any telco experience that can help? I have a "line simulator" in one of the labs at work that should be able to do it, but it's probably a huge investment - serious overkill. I looked at the UK link earlier, and that is more complicated that I want to mess with at this point - I'm looking for a solution that will not require me to breadboard - idealy something cheap and off the shelf.


Lastly, any of the Replay guys want to get involved? I *love* my toy, and would rather have this as a blessed solution - or at least not create too much trouble...



Regards,

jld


ps. Anyone who uses w2k pro as a RAS server, drop me a line - I'd like to see if you can duplicate this without Linux.
 

·
Registered
Joined
·
1,708 Posts
Just a bump and a request whether any progress has been made here. If I had the time I would be working on it too.


Scyber
 

·
Registered
Joined
·
3,715 Posts
Quote:
Originally posted by seanriddle:
Just got back from Radio Shack. I've got a shoebox of resistors, but no 560 ohm 1/2 W.... I'll keep you posted.
Any luck yet, Sean?


-Gary
 

·
Registered
Joined
·
339 Posts
Quote:
Originally posted by seanriddle:
I'm going to order the ring down unit from Jameco so I don't sit idle trying to get this circuit to work. If I do get it working, I'll post the Jameco unit for sale here.
You might consider ordering it direct from the http://www.camblab.com folks. There seem to be 3 different units and the middle one might be best. Read their pdf files to learn more about the differences.

 

·
Registered
Joined
·
912 Posts
Quote:
Originally posted by seanriddle:
Yeah, I actually found that one yesterday when I ran into a snag. There are a couple of problems- it does a lot more (and therefore is much more complicated) than needed for this project, and it was designed for the UK. The difficulty there is that ringing voltage is brought in on a seperate 3rd wire, unlike the US where it is AC superimposed on the DC battery voltage.
The 3rd wire is just cap-coupled to the 2nd wire to provide the separate ring signal. If you just leave off the 3rd wire that circuit should work in the US as well.

Quote:


Can any EEs out there give me a hand? I read that you can detect off-hook by the changing impedence, but I don't know how to build that circuit. The UK circuit detects the increased current when the phone goes off-hook- is that a good way to do things?
Phone lines are typically 20V with no load (this is from memory so these numbers probably aren't quite right, but the concept still holds). When off-hook they place the equivalent of a 1kohm load, and the voltage drops to 6V. The following circuit should be able to detect this change.


The 12V supply with the 1k resistor would be the voltage supply for the phone circuit. When on-hook (no load on the phone line) the 12V supply will turn on the transistor. When off-hook (phone line gives 1k load) the line voltage drops to 6V (below the zener voltage), and the transistor turns off.

Code:
Code:
12V
     ---
      |
      <
      >  1k
      <                       o  Pick-up detect
      |                       |
      |   100k   7V zener    |/
 o----+---/\\/\\/----|<|---+---|  NPN
                         |   |\\
phone                    >    |
line                  1M <    |
                         |    |
 o-----------------------+----+
 

·
Registered
Joined
·
912 Posts
Quote:
Originally posted by seanriddle:
I can detect off-hook when I'm not ringing the phone, using a couple of zener diodes. When I am ringing the phone, I send ~60VAC to it, so I use a relay to switch out my circuit and switch in the AC. That's where my problem is, since now my circuit isn't there to detect off-hook- and if it was still connected, wouldn't the AC screw things up?
OK, I see where you are going. Since you can detect the off-hook condition when not ringing, the problem is that you need to stop the ring signal immediately after the phone goes off-hook, instead of waiting for the 2 second ring to complete. This makes sense both because you want a quick answer, and you don't want to drive an off-hook device with a 60V ring signal.


If you make a circuit that senses an increase in current during the ring pulse, and use this signal to immediately turn off the ring pulse relay, then your other detection circuit will be able to detect the off-hook condition and complete the connection cycle.


You can pass the 50VAC source through a resistor before it connects to the phone line. Then half-wave rectify the voltage on the phone line to get the peak value, and compare that peak value against a threshold. When the ringing phone is on-hook (high impedance) the peak value will be 60V. When the ringing phone is off-hook (1kohm impedance) then the peak value will be about half of that, depending on your choice of resistors.

Code:
Code:
+------------o to phone line through ring relay
              |
              |                                      o off-hook detect
              |                 30V                  |
        1k    |  diode         zener               |/
o------/\\/\\/--+--|>|-----+----+--|<|---/\\/\\/--+----|  NPN
50VAC                    |    |               |    |\\
ring                    ===   <               <      |
signal                   |    >               >      |
                         |    |               |      |
o------------------------+----+---------------+------+
If the peak voltage on the phone line is >30V (zener value) then the transistor will turn on. As soon as the phone line picks up and draws enough current to drop the voltage below 30V, the transistor will turn off.


Tie the off-hook detect to whatever you have enabling the ring relay so that when this transistor turns off, the ring relay cycle gets reset to the between-rings state.
 

·
Registered
Joined
·
75 Posts
Just got my hands on both the Standard and Economy Cambridge units, and they work great. The primary difference between the two is the ability to use a DC power source on the Standard unit. The economy unit also does not have integrated RJ11 connectors on it, just screw-downs. I'll probably end up keeping the economy unit - thanks to RJS for the evals!


That said, assuming that you had your replay connected to a pc via the Cambridge unit, you now have the ability to ring the modem on the PC when the Replay goes off-hook.


I configured a Linux box to act as a PPP server, and added the account name and password that are used by my unit to connect to Replay. This linux box is connected to a "private network" that lives on my side of a cable modem. I serve a DHCP address, with DNS and Gateway info to the PPP connection.


This configuration seems to work perfectly - the only hitch is that I am unable to duplicate this with W2Kpro due to account name issues - as SR pointed out earlier, the account name Replay uses is one character too long to use with windows - there is a workaround on W2Kserver with IAS, but that is a extreme solution IMHO. Hacking your replay to change this works, but then any hacking on your replay is "At your own risk" ;-)


If anyone knows of PPP server software that runs on Windows but allows for non-windows authentication, let me know - I'd be interested in seeing if it would work.


SO - if you are willing to spend the $80US on the Cambridge hardware, and setup a Linux box as a ppp server, you can ditch your phone connection and avoid LD charges.

 
1 - 20 of 173 Posts
Status
Not open for further replies.
Top